Nirmatrelvir-Ritonavir Long COVID Treatment Shows No Benefit in RECOVER-VITAL

Nirmatrelvir-ritonavir long COVID treatment shows no benefit, according to results from RECOVER-VITAL, the largest randomised trial of the antiviral approach to date. The phase 2 study, published in The Lancet Infectious Diseases, found no improvement across three distinct symptom phenotypes and leaves the viral persistence hypothesis of long COVID without a proven therapy.

Investigators tested whether nirmatrelvir-ritonavir, the antiviral combination sold as Paxlovid, could ease symptoms tied to cognitive dysfunction, autonomic dysfunction, and exercise intolerance. Between July 27, 2023, and September 6, 2024, the trial screened 1,207 people and randomised 964 across 69 US sites, with 959 participants included in the analysis: 332 with cognitive symptoms, 334 with autonomic symptoms, and 332 with exercise intolerance.

Nirmatrelvir-ritonavir long COVID: no benefit across three phenotypes

Participants received either 15 days of nirmatrelvir-ritonavir followed by 10 days of placebo, 25 days of nirmatrelvir-ritonavir, or 25 days of placebo-ritonavir. The primary endpoint was a clinically meaningful change in a phenotype-specific patient-reported outcome measure at day 90.

Ad

No comparison reached statistical significance. For cognitive symptoms, the 25-day regimen improved outcomes by 3.2 percentage points versus placebo (95% CI -10.4 to 16.8, p=0.65), and the 15-day regimen by -2.2 percentage points (-15.5 to 11.1, p=0.74). For autonomic symptoms, the adjusted differences were -6.4 percentage points (-18.5 to 5.7, p=0.30) and -0.1 percentage points (-12.5 to 12.3, p=0.99). For exercise intolerance, they were -7.8 percentage points (-19.5 to 3.8, p=0.19) and 0.9 percentage points (-11.4 to 13.2, p=0.88).

Secondary performance measures told the same story, and subgroup analyses by vaccination status, time since infection, race, ethnicity, sex, and initial COVID-19 treatment showed no effect.

A placebo response that demands caution

The most instructive finding may be what happened in the placebo group. Large shares of participants who received only low-dose ritonavir improved over time: 55% met the cognitive improvement threshold, 70% met the autonomic threshold, and 33% met the exercise threshold. This background rate exceeded the assumptions used to calculate the sample size, and the authors attribute it to the Hawthorne effect and the natural variability of long COVID symptoms rather than any activity of ritonavir.

That pattern matters for evidence assessment. It means uncontrolled observations of improvement, including anecdotal reports that antivirals help long COVID, should be read with caution, and it reinforces why placebo-controlled designs are essential before a treatment can support a reimbursement or access case.

Safety and tolerability

Treatment was generally well tolerated, with no new safety concerns from extending the usual 5-day course to 25 days. There were no deaths, and 52 serious adverse events occurred in 42 participants, around 4% of the safety population. Diarrhoea and nausea were more common with nirmatrelvir-ritonavir than with placebo: diarrhoea affected 56 (17%) of participants on the 25-day regimen, 58 (18%) on the 15-day regimen, and 38 (12%) on placebo, while nausea affected 37 (11%), 44 (14%), and 23 (7%).

What this means for long COVID research

The negative result does not close the book on viral persistence. The trial did not restrict enrolment to people with evidence of persistent virus, so whether an antiviral could help that subgroup remains open. Blood samples were collected to test for persistent virus and whether viral antigen clears with treatment, and those analyses are pending.

The authors argue the field needs better tools to measure long COVID symptom burden and a clearer picture of its underlying biology. For a repurposed antiviral, the result also leaves little basis for expanded use in long COVID. Future trials will likely need biomarker screening at baseline to identify participants most likely to respond, and possibly longer treatment courses or combination antiviral approaches.

The study was funded by the National Institutes of Health as part of the RECOVER initiative, with study medication donated by Pfizer. The trial is registered at ClinicalTrials.gov (NCT05595369). Source: The Lancet Infectious Diseases.